J. C. v. Bluntschli, Politik als Wissenschaft. Stuttgart, 1876. 1 vol. 8vo.

For administrative science in particular:

Theod. D. Woolsey, Political Science; or, The State Theoretically and Practically Considered. London, 1878, 2 vols. 8vo.

L. v. Stein, Die Verwaltungslehre, 1865 et seq., 7 vols. 8vo. This is a very important work of colossal size, but incomplete and not free from errors. The author most usefully completed and abridged his work in his Handbuch der Verwaltungslehre.

1876. 1 vol. 8vo. (2nd ed).

Among Italian writers we must not forget:

Stuttgart,

G. Manna, Partizioni teoretiche del diritto amministrativo. 2nd ed. Naples, 1860.

C. F. Ferraris, Saggi d'economia, di statistica e di scienza dell' amministrazione.

learned and remarkable essays).

Turin, 1880. (Very

CHAPTER IV.

ON THE METHOD OF POLITICAL ECONOMY.

§ 1.

By method we mean the logical process by which truth is discovered (i.e. the inventive method) and by which it is communicated to others (i.e. the demonstrative or didactic method).

The study of the inventive method (method in its strict sense) is of the first importance, not only in philosophy but also in the special sciences, where it is often neglected. The didactic method has a purely educational value.

Notwithstanding that the general doctrine of method belongs to the sphere of logic we will touch on it briefly here, in order to render intelligible the many lively controversies which are constantly going on about the method best suited to political economy. And first to explain the distinction between the deductive and inductive methods, which constitute two poles logically opposed.

In the deductive method (synthetic, rational, à priori), certain general principles being given which are self-evident or known by immediate intuition, the

D

consequences comprehended in them are deduced by a connected chain of reasonings. It thus proceeds from the general to the particular by the help of pure reason, and with no external aid. It is an exact method, leading to certain results, provided that the premisses are true and sufficient, and the deductions correct.

The inductive method (analytical, empirical, à posteriori) is founded on the observation of certain phenomena; and by means of abstractions stating their analogies and differences, it seeks to discover the laws by which they are governed. It proceeds therefore from the particular to the general with the help of what is called inductive reasoning-reasoning, that is, which is based on experience. It is a less rigorous method than the deductive, and leads to results which in relation to truth are merely more or less probable, it not being possible to arrive by it at absolute certainty.

According to the different uses of these methods we have deductive sciences, inductive sciences, and sciences which are partly deductive, and partly inductive.

The purely deductive or exact sciences in the strict sense are specially the mathematical sciences, and in part only the philosophical sciences. Thus they include arithmetic, algebraic analysis, geometry, and in fact every branch of abstract mathematics.

The exactness of mathematics is owing firstly to the use of deductive reasoning; and secondarily to the use of a symbolic language (ciphers, letters, geometrical figures). This is superior to common language, both from its brevity and elegance and because it indicates all the logical operations which have been performed in the reasoning, thus facilitating the discovery and the

correction of such errors as may have marred it. And here we must beware of confounding the accessory (the symbols) with the principal, since it is quite possible to apply ordinary language to mathematics, though we thus lose in brevity and elegance.

Purely inductive sciences are to be found among the physical sciences, which, having until now remained. in a purely descriptive stage, are consequently little more than mere generalisations of phenomena. Such, for example, are geology and meteorology, notwithstanding their recent notable advances. Among social sciences, we have statistics, which, although it makes use of what is now sometimes called mathematical induction, has not yet passed beyond this first stage.

Sciences partly inductive and partly deductive are those which, having found their general principles by means of induction, proceed deductively to discover the consequences of these principles, with the frequent use of calculation. The more advanced physical and natural sciences belong to this group, as for example chemistry, physics, mechanics and astronomy. Such sciences, inductive in their beginnings but subsequently reaching the deductive stage, may in a wide sense be called exact sciences, in addition to abstract mathematics, which is, as has been seen, purely deductive.

But here we ought to make another distinction, important in itself, and also because it helps people to avoid errors into which they may fall if they follow literally certain nomenclatures, frequent enough in common usage, but somewhat incorrect.

Those sciences which are founded on observation and induction, and which some authors loosely and vaguely call experimental sciences, are really divided into two

classes according to their very different modes of

observation.

These are:

1. Sciences of observation (in the strict sense), which study certain phenomena as they naturally present themselves, without subjecting them to any modification. And among these again those sciences in which observations are only made by means of the unaided senses are to be distinguished from those in which it is possible to make use of instruments of precision, as astronomy, meteorology, &c.

2. Experimental sciences, which unite with simple observation the use of experiments properly so called, consisting in the artificial production of phenomena with a constant varying of conditions. So, for example, physics and chemistry owe the greater part of their progress to the possibility of experiments, in other words observation carried to a greater degree of perfection.

§ 2.

The method proper to the social sciences, these being as a whole younger and less advanced than the physical or mathematical sciences, is still a much controverted point, and more or less profound disputes still take place, on their logical character, on their nature as deductive or inductive sciences, and on the possibility of applying experimental investigation to them. These disputes rarely lead to any results, being often raised without sufficient philosophic preparation, and especially by dilettanti, who at the least excuse digress from the field of letters into that of the social sciences which they lightly consider as belonging to them.